Bringing Human-Like Reasoning to Driverless Car Navigation

by Rob Matheson, MIT News

 

With aims of bringing more human-like reasoning to autonomous vehicles, MIT researchers have created a system that uses only simple maps and visual data to enable driverless cars to navigate routes in new, complex environments.

Human drivers are exceptionally good at navigating roads they haven’t driven on before, using observation and simple tools. We simply match what we see around us to what we see on our GPS devices to determine where we are and where we need to go. Driverless cars, however, struggle with this basic reasoning. In every new area, the cars must first map and analyze all the new roads, which is very time consuming. The systems also rely on complex maps — usually generated by 3-D scans — which are computationally intensive to generate and process on the fly.

Massachusetts Institute of Technology (MIT) researchers have invented a system to enable autonomous vehicles to navigate complex environments by checking a basic global positioning system-like map and employing video feeds from cameras. The autonomous control system first “learns” humans’ steering patterns on suburban streets via the map/video combination.

In training, a convolutional neural network correlates steering wheel spins to road curvatures, as seen through cameras and the inputted map. Once trained, the system can pilot a car along a preplanned route in a new area by emulating a human driver.

The system also continuously identifies any mismatches between the map and road features to sense whether its position, sensors, or mapping are wrong, and makes appropriate course corrections. Said MIT’s Alexander Amini, “With our system, you don’t need to train on every road beforehand. You can download a new map for the car to navigate through roads it has never seen before.”

“Our objective is to achieve autonomous navigation that is robust for driving in new environments,” adds co-author Daniela Rus, director of the Computer Science and Artificial Intelligence Laboratory (CSAIL) and the Andrew and Erna Viterbi Professor of Electrical Engineering and Computer Science. “For example, if we train an autonomous vehicle to drive in an urban setting such as the streets of Cambridge, the system should also be able to drive smoothly in the woods, even if that is an environment it has never seen before.” Read the full report

 

DCL: If you believe this is going to work anytime soon, I’ve a bridge in Brooklyn you should buy. I’ve watched “driverless cars” trying to navigate the streets in my neighborhood and they are totally incapable of dealing with situations like overtaking large vehicles on four lane roads (two lanes in each direction) with no center divider. They are very likely to cause accidents.